An intelligent moving window sparse principal component analysis-based case based reasoning for fault diagnosis: Case of the drilling process. (September 2022)

• Record Type: Journal Article
• Authors: Han, Yongming; Liu, Jintao; Liu, Fenfen; Geng, Zhiqiang
• Is Part Of: ISA transactions. Volume 128(2022)Part A
• Page Start: 242
• Page End: 254
• Publication Date: 2022-09
• Subjects: Fault diagnosis -- Moving window sparse principal component analysis -- Case-based reasoning -- Drilling process -- Petrochemical industry; Engineering instruments -- Periodicals; Engineering instruments; Periodicals; Electronic journals; 629.805
• DOI: 10.1016/j.isatra.2021.09.016
• Languages: English
• ISSNs: 0019-0578

Abstract

Abstract: The drilling process is an important step in petrochemical industries, but the drilling process is risky and costly. In order to improve the safety and cost the impact of faults in the drilling process, this paper proposes intelligent moving window based sparse principal component analysis (MWSPCA) integrating case-based reasoning (CBR) (MWSPCA-CBR) in the fault diagnosis of the drilling process in the petrochemical industry. Through introducing sparsity into the PCA model, the Lasso constraint function of the MWSPCA method is used to optimize the sparse principals. The corresponding T 2 and Q statistics calculated by the selected sparse principals decide whether the faults have occurred, and the occurrence time of the anomaly is quickly located based on the MWSPCA method. Then the CBR method is used to analyze the anomaly data to identify the possible fault types, and provide the relational handling methods for real-time monitoring experts. Finally, the MWSPCA method is verified based on the intelligent diagnosis of the Tennessee Eastman (TE) process, reducing false negatives and false positives and improving the accuracy rate and the diagnosis speed. Furthermore, the proposed method is applied to analyze the data of the drilling process. The experimental results demonstrate that the proposed method can effectively diagnosis faults in the drilling process and reduce risks and costs in the petrochemical industry. Graphical abstract: Highlights: An improved moving window sparse principal component analysis-based case based reasoning is proposed. The moving window sparse principal component method is verified based on the diagnosis of the Tennessee Eastman process. The intelligent fault diagnosis framework of the drilling process is obtained. The proposed method can effectively diagnosis faults in the drilling process. The proposed method can reduce risks and costs of petro chemical engineering.